The art of AI engineering — harnessing the power of AI for reliable use — Part I
AI engineering is building reliable products on top of pre-trained models: knowing when to hire the brain, and evaluating it honestly.
- Date
- August 2026
- Category
- AI
Berlin, Germany
Product manager // AI engineer.
I build AI products end to end, from product requirements and planning to evals, harness, observability, AI security, deployment and EU compliance. In my reflection essays, I distil what I learn into frameworks that optimise the process.
Start here · seven short decks
The art of AI engineering, Part I, in 14 slides
Read the full essay →
01 / 14
Download slides (PDF)My essays about Tech & the Future
→The art of AI engineering — harnessing the power of AI for reliable use — Part II
The model is one piece. The scaffolding around it, plus security, observability and continuous improvement, is what makes it reliable in production.
- Date
- September 2026
- Category
- AI
Case study: AI-powered travel RAG system — what I learned from AI engineering
Planned for 16 weeks, public in three: a ruthless MVP, what was cut or deferred and why, and why the model was the easiest part.
- Date
- September 2026
- Category
- AI
Traditional ML in practice: predicting follow-on funding from Crunchbase data
A step-by-step ML project: predicting which startups raise again within two years, from EDA and label design to a model whose top decile raises again at 69%.
- Date
- September 2026
- Category
- AI
The importance of our climate in history of human civilisation
Climate has shaped where and how human civilisation grew, which is why a changing climate matters so much for our future.
- Date
- July 2026
- Category
- Climate change
The search for clean energy
Humanity has been decarbonising its energy for centuries. Nuclear, renewables and hydrogen are how we could finish the job.
- Date
- July 2026
- Category
- Climate change